← BACK TO AI LAB

AI LAB / Agents / Evaluation

Agent Tools Under Failure

Testing agent recovery when tools time out, reject input, duplicate effects, or return partial data.

AgentsEvaluationReliability
EXPERIMENT
EXP-015
STATUS
complete
DATE
Aug 21, 2026
TAGS
Agents · Evaluation · Reliability

EXPERIMENT / HYPOTHESIS

Explicit typed failure states will outperform natural-language error messages in multi-step agent recovery.

  1. 01Create 120 deterministic failure scenarios
  2. 02Run identical tasks with typed and prose-only errors
  3. 03Score completion, unsafe retries, and wasted calls
Pass rate
86.4%
Unsafe retries
2.1%
Tests
120

CONCLUSIONTyped failures improved successful recovery and sharply reduced repeated side effects. The remaining failures were mostly planning mistakes rather than tool handling.

01

Failure matrix

Scenarios covered validation, authorization, timeout-before-effect, timeout-after-effect, partial result, and dependency outage.

  • Retryable
  • Terminal
  • Unknown outcome
02

Finding

The agent recovered best when the orchestrator supplied a small stable error code, retry guidance, and confirmed operation state.

  • Less prompt ambiguity
  • Safer retries
  • Easier evaluation